A Multi-Frequency PWM Scheme for Multi-Level

Five-Phase Open-End Winding Drives

Wahyu Satiawan, Martin Jones
Liverpool John Moores University, School of Engineering, Liverpool, UK
Email: i.n.satiawan@2008.ljmu.ac.uk
Abstract This paper presents a multi-frequency pulse width
modulation (PWM) scheme for a multi-level five-phase open-end
winding drive using the unified modulation technique. The open-
end winding machine is supplied via a two-level voltage source
inverter from each side of the windings. One inverter operates in
ten-step mode while the other is operated under PWM. The
multi-frequency PWM scheme is aimed to control the magnitude
of the fundamental output and to cancel low-order harmonics
produced by the inverter operating in ten-step mode. Simulation
results show excellent performance resulting in multi-level
output voltages with a low THD. Implementation of the scheme
is straightforward, compared with the space-vector modulation
based approach, since space vector transformation, sector
identification, and look up tables are not required.
Index Terms dual-inverter supply, open-end winding
machine, multi-frequency pulse width modulation, unified
modulation technique.
I. INTRODUCTION
Multi-level and multi-phase voltage source inverters (VSI)
have been attracting increasing research interest recently due
to their ability to overcome voltage and current limitations of
power semiconductors and their inherent ability to tolerate
faults [1]-[2]. There are several configurations of multi-level
converters, the main ones being the neutral point clamped
(NPC), the flying capacitor (FC) and cascaded converters [2-
5].
Among the cascaded converters, the dual two-level inverter
configuration has received growing attention due to its simple
structure. The additional diodes used in the diode-clamped
(NPC) VSI are not needed, leading to a saving in the overall
number of components. Furthermore, the issue of proper
capacitor voltage balancing does not exist if the supply is
two-level at each winding side. Typically, three-phase VSIs
are utilised. Application of such a dual-inverter supply
enables drive operation with voltage waveform equivalent to
the one obtainable with a three-level VSI in single-sided
supply mode [5]. Such drives are being considered as
alternative supply solutions in EVs/HEVs [6]-[7] and electric
ship propulsion [8].
Due to their well known advantages [1], multi-phase drives
have also been considered for similar applications as the
multi-level drives. The only known examples where multi-
phase drive systems in open-end winding configuration have
been developed are [9-11] where an asymmetrical six-phase
induction motor drive was considered in [9]-[10].
This paper proposes a new, and relatively simple, PWM
algorithm for the five-phase open-end winding three-level
drive supplied by two two-level five-phase inverters. The
PWM algorithm is based on the decomposition of the there-
level space vector decagon into a number of two-level
decagons. A similar idea has been proposed in [12] for a
three-phase NPC inverter and in [13] for a three-phase open-
end winding drive. In the case of the five-phase open-end
winding drive the situation is significantly more complicated
since both 2-D planes have to be considered. It is shown that
the proposed modulation strategy is capable of achieving the
target fundamental while eliminating any low-order harmonic
content in the output load phase voltage.
The paper begins with a review of the five-phase two-level
drive characteristics, which is followed by a general
description and mathematical model of the open-end winding
topology, along with mapping of the space vectors into the 2-
D planes. Next, the proposed modulation method is described.
It is shown that due to the nature of the five-phase topology
one of the inverters needs to operate with so called multi-
frequency PWM. Finally, the performance of the proposed
modulation method is verified via simulation.
II. GENERAL PROPERTIES OF TWO-LEVEL FIVE-PHASE
DRIVES
Although the proposed modulation technique can be
considered a carrier based PWM method the space-vector
approach is an extremely useful tool when trying to
understand the relationships which govern the performance of
five-phase drives and the corresponding VSI modulation
techniques. A five-phase machine can be modelled in two 2-
D sub-spaces, so-called - and x-y sub-spaces [14]. It can be
shown that only current harmonic components which map
into the - sub-space develop useful torque and torque
ripple, whereas those that map into the x-y sub-space do not
contribute to the torque at all. A multi-phase machine with
near-sinusoidal magneto-motive force distribution presents
extremely low impedance to all non-flux/torque producing
supply harmonics and it is therefore mandatory that the
supply does not generate such harmonics. What this means is
that the design of a five-phase PWM strategy must consider
simultaneously both 2-D sub-spaces, where the reference
voltage, assuming pure sinusoidal references, is in the first
plane while reference in the other plane is zero. Two-level
five-phase inverters can generate up to 2
5
=32 voltage space
vectors with corresponding components in the - and x-y
sub-spaces. Active (non-zero) space vectors belong to three
groups in accordance with their magnitudes - small, medium
and large space vector groups. The magnitudes are identified
UPEC 2011 46th International Universities' Power Engineering Conference 5-8th September 2011 Soest Germany
ISBN 978-3-8007-3402-3 VDE VERLAG GMBH Berlin Offenbach
with indices s, m, and l and are given as
dc s
V v ) 5 2 ( cos 5 4 t = ,
dc m
V v 5 2 = , and
dc l
V v ) 5 ( cos 5 4 t = , respectively. Four active space vectors
are required to generate sinusoidal voltages [1]. In order to
provide zero average voltage in the x-y plane two
neighbouring large and two medium space vectors are
selected [14]. It is shown in [14] that the maximum peak
value of the output fundamental phase-to-neutral voltage in
the linear modulation region is v
max
= 0.525V
dc
, resulting in a
maximum modulation index, M =1.05.
III. FIVE-PHASE OPEN-END WINDING TOPOLOGY
Fig. 1 illustrates the open-end winding structure, based on
utilisation of two two-level five-phase VSIs. The two
inverters are identified with indices 1 and 2. Inverter legs are
denoted with capital letters, A,B,C,D,E and the negative rails
of the two dc links are identified as N1 and N2. Machine
phases are labelled as a,b,c,d,e. Phase voltage positive
direction is with reference to the left inverter (inverter 1).
Two isolated dc supplies are assumed so that the common
mode voltage (CMV)
2 1N N
v is of non-zero value (the issue of
CMV elimination is not addressed here). The resulting space
vectors in dual-inverter supply mode will depend on the ratio
of the two dc link voltages. The situation considered further
on is the setting 2 /
2 1 dc dc dc
V V V = = , which gives the
equivalent of single-sided three-level supply. Using the
notation of Fig. 1, phase voltages of the stator winding can be
given as [11]:
2 2 2 1 1 1
2 2 2 1 1 1
2 2 2 1 1 1
2 2 2 1 1 1
2 2 2 1 1 1
N E N N N E es
N D N N N D ds
N C N N N C cs
N B N N N B bs
N A N N N A as
v v v v
v v v v
v v v v
v v v v
v v v v
+ =
+ =
+ =
+ =
+ =
(1)
Space vectors of phase voltages in the two planes are
determined with;
( )( )
( )( )
e d c b a
y x
e d c b a
v a v a v a v a v v
v a v a v a v a v v
8 6 4 2
4 3 2
5 2
5 2
+ + + + =
+ + + + =

| o
(2)
where ) 5 / 2 exp( t j a = . Using (1) and (2), one gets
) 2 2 2 2 2 ( ) 1 1 1 1 1 (
) 2 2 2 2 2 ( ) 1 1 1 1 1 (
E D C B A y x E D C B A y x y x
E D C B A E D C B A
v v v
v v v


=
= | o | o | o
(3)
since 0 ) 1 (
4 3 2
2 1
= + + + + a a a a v
N N
. In (3) the two space
vectors on the right-hand sides of the two equations are
corresponding voltage space vectors of the two five-phase
two-level VSIs. There are three vector lengths, large, medium
and small: 0.32366V
dc
, 0.2V
dc
, 0.123V
dc
. Voltage space
vectors, produced by the 1024 possible switching states, are
illustrated in Fig. 2 for the o| plane. There are 211 space
vectors, 21 active phase voltage space vectors per 36 degrees
sector. Space vector mapping into the x-y plane follows the
pattern that exists for a five-phase VSI, the largest vectors of
the first plane map into the smallest vectors of the second
plane, and vice versa [11].
Fig. 1. Dual-inverter fed five-phase open-end winding topology.
Fig. 2. Voltage space vectors of dual-sided structure decomposed into smaller
decagons.
IV. PRINCIPLE OF THE PROPOSED ALGORITHM
The design of a suitable modulation method for the dual-
fed structure represents a considerable challenge due to the
high number of switching states. An attempt is made here, by
at first decomposing the voltage vectors into smaller decagons
equal to two-level structure, as shown in Fig. 2. There are 10
small decagons in the outer ring (the centre points are labelled
as A, B, C, D, E, F, G, H, I, J) and one decagon in the centre
(centred at point O). The centre decagon comprises of vectors
which can be activated if one inverter is used up to half of the
achievable maximum voltage with the other one short-
circuited. As a consequence, the converter is in two-level
mode of operation and the PWM is based on four active and
zero vector application, as discussed in section II and [14]. As
can be seen in Fig. 2, the origins of the outer decagons are
located on the outer vectors of the inner decagon, denoted
by the black dots in Fig 2. In the case of the three-phase
topology [12] operation in the outer region (outer hexagon) is
achieved when one inverter operates with a single voltage
space vector applied (the nearest one to the reference) and the
second inverter is modulated using the standard three-phase
two-level space vector modulation (SVM) technique. Let the
applied vector for the one inverter be v
i
and let the reference
be v*. Here v
i
is the vector produced by the inverter that is the
nearest to the reference, and v* exceeds, in magnitude,
maximum voltage realisable with one inverter. The reference
for the other inverter is then set as:
UPEC 2011 46th International Universities' Power Engineering Conference 5-8th September 2011 Soest Germany
ISBN 978-3-8007-3402-3 VDE VERLAG GMBH Berlin Offenbach
) * ( * *
i
v v v = (4)
In other words, when the magnitude of the reference
voltage exceeds the maximum value obtainable with one
inverter, one inverter is operated in six-step mode while the
second inverter is modulated in the standard way.
It is well known that operation of a five-phase inverter in
ten-step mode without a controllable dc link leads to
uncontrollable fundamental output voltage magnitude and
unwanted low order harmonics, which for the leg voltage can
be expressed as a Fourier series as follows:
( ) ...] 7 sin
7
1
5 sin
5
1
3 sin
3
1
[sin
1
t t t t V t v
dc
e e e e
t
+ + + = (5)
In a five-phase system, harmonics of the order 10k1 (k =
0, 1, 2, 3...) map into the torque/flux producing subspace, -,
while harmonics of the order 10k3 map into the x-y
subspace. They do not produce any useful torque/flux and
simply lead to large unwanted loss-producing currents. The
large currents are a consequence of the relatively small
impedance presented in x-y plane. 10k5 are zero-sequence
components. This leads to the requirement that the second
inverter must be able to not only control the fundamental but
also eliminate the unwanted low order harmonics, which are
produced by applying only the large vector in o| from one
inverter. This causes unwanted harmonics in both planes
since a large vector has a corresponding non-zero value in the
second, x-y plane [14]. In order to achieve this objective, the
second inverter modulation scheme will need to operate in
both the - and the x-y planes. Such PWM modulators are
known as multi-frequency PWM modulators and were first
developed in order to control multiphase multi-motor drive
systems [15].
The proposed scheme is realised by utilizing two separate
two-level modulators. Fig. 3 shows the voltage vectors that
are used by the modulators. Inverter 1, which operates in ten-
step mode, applies a single vector in one switching period.
The vectors are mapped in the points of the outer ring of
decagon (labelled as A, B, C, D, E, F, G, H, I, J in Fig. 2);
these correspond to the large vectors in a two-level modulator
(Fig. 3(a)). The modulator for inverter 2, which operates in
PWM mode, is able to choose from all 32 voltage vectors
(small, large, medium and zero) and select suitable 4 active
and one zero vectors in-order to control the fundamental and
cancel unwanted harmonics in the output. The multi-
frequency PWM is employed for the higher modulation
indices (0.525 < M 1.05). In the lower modulation index
region (M 0.525) only one inverter operates, which reverts
the system into single-sided two-level SVM operation mode.
Fig. 4 shows the voltage reference waveform (trace (c)) of the
multi-frequency modulator for modulation index of 1.05. It is
obtained by subtracting the ten-step mode inverter phase
voltage output (trace (b)) from the total voltage reference
(trace (a)) and inverting the signal. It can be noticed that the
resulting reference of the PWM modulator is multi-frequency.
The non-sinusoidal reference is difficult to achieve using the
space-vector modulation technique. The control method
developed in the paper adapts the unified PWM technique
|
.
|

\
|
5
cos 2
5
2
1
t
dc V |
.
|

\
|
5
cos 2
5
2
2
t
dc V
Fig. 3. The voltage space vectors employed for ten-step modulator (a) and the
multi-frequency modulator (b).
Fig. 4. Generation of the voltage references for multi-frequency modulator.
[16] to function as a multi-frequency PWM modulator, which
is capable of controlling the fundamental, eliminating
unwanted low-order harmonics and maintaining the
maximum dc link utilisation. The technique is well-
established for three-phase systems and has been recently
extended to the single sided two-level five-phase system, as
reported in [17]. Implementation of the unified method is
advantageous as the computational load is greatly reduced
owing to the removal of the space vector transformation,
sector identification and switching state look-up table.
V. UNIFIED VOLTAGE MODULATION TECHNIQUE FOR
MULTI-FREQUENCY OUTPUT GENERATION
In the conventional SVM scheme, calculation of dwell
time is based on the magnitude of the voltage reference space
vector and its relative angle with respect to the reference axis.
Space vector transformation, sector identification and
information regarding the selected vectors are required to
perform the task. In the unified PWM technique the dwell
time is determined directly from instantaneous values of the
phase voltage reference. This is so since the effective time
evaluated in the conventional SVM is just the difference
value between two application times corresponding to the
phase voltage [16], [17]. The basic principle of the method is
shown in Figs. 5 and 6.
In what follows it is assumed that the inverter 2 output is
modulated. Regardless of the sector where the reference
vector is, the application time for each phase voltage can be
defined using:
0
0
* * * * * * * * * *
= + + + +
= + + + +
es ds cs bs as
es ds cs bs as
T T T T T
v v v v v
(6)
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02
-400
-300
-200
-100
0
100
200
300
400
References for modulation index, M = 1.05
(
V
)
Time (s)
(a)
(c)
(b)
UPEC 2011 46th International Universities' Power Engineering Conference 5-8th September 2011 Soest Germany
ISBN 978-3-8007-3402-3 VDE VERLAG GMBH Berlin Offenbach
* *
2
* *
2
* *
2
* *
2
* *
2
,
, , ,
es
dc
s
es ds
dc
s
ds
cs
dc
s
cs bs
dc
s
bs as
dc
s
as
v
V
T
T v
V
T
T
v
V
T
T v
V
T
T v
V
T
T


(7)
where T
s
is the switching period. The application times, T
as
,
T
bs
, T
cs
, T
ds
and T
es
which are then named as imaginary
switching times are proportional to the instantaneous phase
voltage reference magnitude. The effective time, T
eff
is
defined as the difference between the maximum and
minimum values among T
as
, T
bs
, T
cs
, T
ds
and T
es
. The effective
times are the time durations of the active vector application.
The time duration of zero vector, T
zero
is obtained by
subtracting from T
s
the effective time [16]:

} , , , , min{
} , , , , max{
min
max
min max
eff s zero
es ds cs bs as
es ds cs bs as
eff
T T T
T T T T T T
T T T T T T
T T T
=
=
=
=
(8)
In order to obtain a symmetrical switching pattern over
switching period, the zero voltage has to be distributed
equally. The offset time T
offset
is required:
min
) 2 / ( T T T
zero offset
= (9)
The actual switching times for each inverter leg can be
obtained by the time shifting operation as follows:
offset es ge
offset ds gd offset cs gc
offset bs gb offset as ga
T T T
T T T T T T
T T T T T T
+ =
+ = + =
+ = + =
,
,
(10)
VI. SIMULATION VERIFICATION
In order to verify the drives performance a series of
simulations were performed. The dc link voltage of each
inverter is 300 V ( 2 /
2 1 dc dc dc
V V V = = ), giving an effective
dc link voltage of 600 V when M > 0.525. The switching
frequency of the modulated inverter is 2 kHz. Ideal operation
of the inverters is assumed, meaning that effects due to the
dead-time and semiconductor voltage drops were neglected.
Fig. 7 shows the load phase voltage and the inverters leg
voltage waveforms and spectra at the boundary between
regions where the PWM modulated inverter reduces the
fundamental voltage, created by the ten-step inverter, or adds
to it (this takes place when M = 0.6366). The converter
achieves the reference fundamental with minimum low order
harmonics. The leg voltages of each inverter show that the
harmonics created by the ten step inverter are counter
balanced by the modulated inverter and the fundamental is
overwhelmingly provided by the ten step modulator. This
means that the modulated inverter is only compensating the
unwanted harmonics. In order to examine the drives
performance over the complete linear operating range
simulations were performed with modulation index of 0.1
1.05 in 0.05 increments. The THD of the phase output voltage
and the respective and x-component is shown in Fig. 8. It
can be seen that the THD remains low for the entire linear

Ts
Phasea
Tas
t1
Phaseb
Tbs
t2
Phasec Vdc/2
Tcs
t3
Phased
Tds
t4
Phasee
Tes
Teff
Tmin Tmax
Fig. 5. Illustration of the imaginary switching times.
Fig. 6. Actual switching pulse pattern.
operating region. The performance is significantly improved
when compared to the single-sided two-level drive in line
with the higher number of phase voltage levels available
particularly in the higher modulation index.
Next, the phase variable model of the five-phase machine
is used and the drive is operated in open-loop V/f mode. The
voltage reference profile is such that the supply frequency of
the machine is ramped from zero to 50 Hz in 0.5s. At the
operating frequency of 50 Hz the maximum modulation index
is reached (M = 1.05). Voltage boost is not applied. The
acceleration transient is presented in Figs. 9 and 10 for 25 Hz
(M = 0.525) and 50 Hz (M = 1.05), respectively along with
the steady-state load phase voltage waveform and spectrum. It
can be seen in Fig. 9 that no voltage reference is applied to
inverter 2 when M = 0.525 and so the drive operates in two-
level mode. For this reason the steady-state phase voltage
contains only nine levels. However, the spectrum is much
improved when compared with its single sided two-level
equivalent due to the 50% reduction in the effective dc link
voltage. Fig. 10 shows that, as soon as M > 0.525 (at t = 0.25
s) inverter 1 operates in ten-step mode, while inverter 2
operates in multi-frequency PWM mode. At t = 0.5 s the
reference frequency is reached and the modulation index is
maximum, M = 1.05. The drive is operating in multi-level
mode and the steady-state phase voltage now comprises 15
levels. The torque ripple is mininal and the stator current is
sinusiodal without any low order harmonics. The load phase
voltage spectra indicate absence of unwanted low-order
Ts Ts
A+gating
Tga
B+gating
Tgb
C+gating
Tgc
D+gating
Tgd
E+gating
Tge Teff
t31/2 to/2
UPEC 2011 46th International Universities' Power Engineering Conference 5-8th September 2011 Soest Germany
ISBN 978-3-8007-3402-3 VDE VERLAG GMBH Berlin Offenbach
harmonics and the target fundamental voltages have been met
in both cases.
VII. CONCLUSION
A multi-frequency PWM scheme for multi-level dual-
inverter fed open-end winding drives has been successfully
Fig. 7. Stator phase and inverter leg voltage waveforms and spectra for M =
0.636, f = 30.31Hz.
Fig. 8. Total harmonic distortion (THD) of the stator phase voltage and the -
axis and x-axis components.
developed based on the unified PWM technique. When M <
0.525, the converter operates in two-level mode utilising a
single inverter. When the reference fundamental exceeds the
capabilities of a single inverter (M>0.525), one inverter is in
ten-step mode and the other is multi-frequency modulated. As
Fig. 9. Motor acceleration under V/f control (ref. 25 Hz, M = 0.525): inverter
modulation signals, speed, torque, stator current and steady-state stator phase
voltage waveform with spectrum.
0.74 0.75 0.76 0.77 0.78 0.79 0.8
-400
-200
0
200
400
fund = 190.6587(pk) THD = 0.58257 levels = 13
(
V
)
Time (s)
0 100 200 300 400 500 600 700 800 900 1000
0
50
100
150
(
V

r
m
s
)
0.74 0.75 0.76 0.77 0.78 0.79 0.8
-200
-100
0
100
200
Inverter 1 leg "A", Mod index =0.63659
(
V
)
Time (s)
0 100 200 300 400 500 600 700 800 900 1000
0
50
100
150
(
V

r
m
s
)
0.74 0.75 0.76 0.77 0.78 0.79 0.8
-200
-100
0
100
200
Inverter 2 leg "A", Mod index =0.63659
(
V
)
Time (s)
0 100 200 300 400 500 600 700 800 900 1000
0
20
40
60
(
V

r
m
s
)
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0
0.5
1
1.5
2
2.5
3
Modulation index
T
H
D
Phase voltage
o-axis
x-axis
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
0
0.5
1
I
n
v
.

1
.

l
e
g

"
a
"

m
o
d
.

s
i
g
n
a
l

(
p
.
u
.
)

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
0
0.5
1
I
n
v
.

2
.

l
e
g

"
a
"

m
o
d
.

s
i
g
n
a
l

(
p
.
u
.
)

Time (s)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
-5
0
5
10
15
T
o
r
q
u
e

(
N
m
)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
0
200
400
600
800
S
p
e
e
d

(
r
p
m
)
Time (s)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
-8
-6
-4
-2
0
2
4
6
8
Time (s)
S
t
a
t
o
r

p
h
a
s
e

"
a
"

c
u
r
r
e
n
t

(
A
)
0.72 0.73 0.74 0.75 0.76 0.77 0.78 0.79 0.8
-300
-200
-100
0
100
200
300
fund = 156.6127(pk) THD = 0.70868 levels = 9
(
V
)
Time (s)
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
0
20
40
60
80
100
120
(
V

r
m
s
)
UPEC 2011 46th International Universities' Power Engineering Conference 5-8th September 2011 Soest Germany
ISBN 978-3-8007-3402-3 VDE VERLAG GMBH Berlin Offenbach
a result the converter is operated in multilevel mode. The
multi-frequency modulated inverter is used to control the
fundamental voltage and eliminate any low order harmonics
created by the ten-step mode inverter. The method has been
verified by simulation.
Fig. 10. Motor acceleration under V/f control (ref. 50Hz, M = 1.05): inverter
modulation signals, speed, torque, stator current and steady-state stator phase
voltage waveform with spectrum.
REFERENCES
[1] E. Levi, Multi-phase electric machines for variable-speed
applications, IEEE Trans. on Ind. Electr., vol. 55, no. 5, 2008, pp.
1893-1909.
[2] P. Lezana, J. Pou, T.A. Meynard, J. Rodriguez, S. Ceballos and F.
Richardeau, "Survey on fault operation on multilevel inverters," IEEE
Trans. on Ind. Electr., vol.57, no.7, 2010, pp.2207-2218.
[3] S. Kouro, M. Malinowski, K. Gopakumar, J. Pou, L. G. Franquelo, B.
Wu, J. Rodriguez, M. A. Perez and J. I. Leon, Recent advances and
industrial applications of multilevel converters, IEEE Trans. on Ind.
Electr., vol. 57, no. 8, 2010, pp. 2553-2580.
[4] S. Lu, S. Mariethoz and K. A. Corzine, Asymmetrical cascade multi-
level converters with noninteger or dynamically changing DC voltage
ratios: Concepts and modulation techniques, IEEE Trans. on Ind.
Appl., vol. 57, no. 7, 2010, pp. 1815-1824.
[5] B. A. Welchko and J. M. Nagashima,A comparative evaluation of
motor drive topologies for low-voltage, high-power EV/HEV
propulsion systems, IEEE Proc. Int. Symp. on Ind. Electronics. ISIE,
Rio de Janerio, Brazil, 2003, pp. 379-384.
[6] C. Rossi, G. Grandi, P. Corbelli and D. Casadei, Generation system for
series hybrid powertrain based on dual two-level inverter, Proc. Eur.
Power Elec. and Appl. Conf. EPE, Barcelona, Spain, 2009, CD-ROM
paper 0978-13043073.
[7] Y. Kawabata, M. Nasu, T. Nomoto, E. C. Ejiogu and T. Kawabata,
High-efficiency and low acoustic noise drive system using open-
winding AC motor and two space-vector-modulated inverters, IEEE
Trans. on Ind. Electr., vol. 49, no. 4, 2002, pp. 783-789.
[8] L. Shuai and K. Corzine, "Multilevel multi-phase propulsion drives," in
IEEE Electric Ship Technologies Symposium, Philadelphia, PA, USA,
2005, pp. 363-370.
[9] G. Grandi, A. Tani, P. Sanjeevikumar and D. Ostojic, Multi-phase
multi-level AC motor drive based on four three-phase two-level
inverters, in Proc. Int. Symposium on Power Electronics., Electrical.
Drives, Automation and Motion, SPEEDAM, Pisa, Italy, 2010, pp.
1768-1775.
[10] K. K. Mohapatra and K. Gopakumar, A novel split phase induction
motor drive without harmonic filters and with linear voltage control for
the full modulation range, EPE Journal, vol. 16, no. 4, 2006, pp. 20-
28.
[11] E. Levi, M. Jones, and W. Satiawan, A multiphase dual-inverter
supplied drive structure for electric and hybrid electric vehicles, in
Proc. IEEE Vehicle Power and Propulsion Conf.,VPPC, Lille, France,
2010, CD-ROM paper 95-45603.
[12] G.Shiny and M.R.Baiju, Space vector PWM scheme without sector
identification for an open-end winding induction motor based 3-level
inverter, Proc. IEEE Ind. Electronics. Soc. Annual Meeting IECON,
Porto, Portugal, 2009, pp. 1310-1315.
[13] E.G. Shivakumar, K. Gopakumar, S.K. Sinha, A. Pittet and V.T.
Ranganathan, Space vector PWM control of dual inverter fed open-
end winding induction motor drive, EPE Journal, vol. 12, no. 1, 2002,
pp. 9-18.
[14] D. Dujic, M. Jones and E. Levi, Generalized space vector PWM for
sinusoidal output voltage generation with multi-phase voltage source
inverters, Int. J. Ind. Electronics. and Drives, vol. 1, no. 1, 2009, pp. 1-
13.
[15] D. Dujic, G. Grandi, M. Jones and E. Levi, A space vector PWM
scheme for multi-frequency output voltage generation with multi-phase
Voltage Source Inverter, IEEE Trans. on Ind. Electr., vol. 55, no. 5,
2008, pp. 1943 1955.
[16] D.W. Chung, J.S. Kim and S.K. Sul, Unified voltage modulation
technique for real-time three-phase power conversion, IEEE Trans. on
Ind. Appl., Vol. 34, no. 2, 1998, pp. 374 380.
[17] A. Iqbal, and M.A. Khan, A simple approach to space vector PWM
signal generation for a five-phase voltage source inverter, Proc. IEEE-
INDICON., Kanpur, India, 2008, vol.2, pp. 418 424.
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
0
0.5
1
I
n
v
.

1
.

l
e
g

"
a
"

m
o
d
.

s
i
g
n
a
l

(
p
.
u
.
)

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
0
0.5
1
I
n
v
.

2
.

l
e
g

"
a
"

m
o
d
.

s
i
g
n
a
l

(
p
.
u
.
)

Time (s)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
-10
0
10
20
T
o
r
q
u
e

(
N
m
)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
0
500
1000
1500
2000
S
p
e
e
d

(
r
p
m
)
Time (s)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
-8
-6
-4
-2
0
2
4
6
8
Time (s)
S
t
a
t
o
r

p
h
a
s
e

"
a
"

c
u
r
r
e
n
t

(
A
)
0.76 0.765 0.77 0.775 0.78 0.785 0.79 0.795 0.8
-500
0
500
fund = 314.6403(pk) THD = 0.32755 levels = 15
(
V
)
Time (s)
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
0
100
200
300
(
V

r
m
s
)
UPEC 2011 46th International Universities' Power Engineering Conference 5-8th September 2011 Soest Germany
ISBN 978-3-8007-3402-3 VDE VERLAG GMBH Berlin Offenbach